AI and Quantum Complexities Shape Security Landscape at RSAC Conference 2025
Cybersecurity Chaos at RSAC 2025
Whew, the RSAC Conference 2025 was a doozy! With cybersecurity experts from around the globe converging on San Francisco, we've been keeping our eyes glued to the action. This year, our platform has been all-inclusive, covering the event from both remote and on-site perspectives.
Topics that took center stage included AI security and the tricks up threat actors' sleeves. In this episode, Jane chats with Rory about the highlights from the conference and the key insights gleaned.
Must-Knows from RSAC 2025
Quantum Cryptography: A Warning from the Future
Quantum cryptography isn't exactly the new kid on the block—it's been a hot topic at RSA conferences for years. And the reason? Simple. It's not a matter of if, but when, quantum computers will become a genuine threat. So whether it's light-years away or just around the corner, it's essential to start preparing.
The Dark Side of AI: Cisco's Warning
Cisco's research shook things up with a sobering dose of reality: when you tweak AI models to better suit your specific organizational needs, you give those models the power to create chaos. Cisco discovered that fine-tuning models can increase the risk of model jailbreaks by up to three times and make them up to 22 times more likely to spew out harmful responses.
Human at Heart
Amidst all the chatter about AI-empowered attackers, it's easy to forget that behind the screens are real people—just like you and me. Despite the sophisticated tools at their disposal, threat actors still make mistakes. They communicate through Discord, Telegram, and other channels, making it possible for human defenders to catch them slipping up.
Two Cents from RSAC 2025*
RSAC Conference 2025 served as a startling reminder of the obstacles facing cybersecurity professionals.RSAC Conference Day One: The mood was all about AI for Security.["Governance is irreplaceable": Microsoft Security VP on the importance of diversity and sector expertise for staying relevant in the era of agentic AI.]RSAC Conference Day Two: The focus shifted towards examining adversarial methodologies.["Exponential effort required": AI security experts called for a renewed focus on the evaluation of innovative models and agentic systems.][Emphasizing the humanity of cyber adversaries, Trellix's research head cautioned that they remain human, just like us.]RSAC Conference Day Three: Exploring the use of AI for more efficiency and facing evolving attack techniques.[China's cyber aggression is on the rise, with organizations slow to catch on to the growing APT threats, according to Kevin Mandia and Nicole Perlroth.]
Join Us*
Stay informed by subscribing to our podcasts on Apple Podcasts or Spotify.Never miss a beat with our newsletter or find us on LinkedIn.
Don't Miss Out*
*Subscribe today and receive a complimentary copy of our Future Focus 2025 report — your go-to guide on AI, cybersecurity, and other IT challenges, as identified by over 700 senior executives.
Keep an eye on our platform for daily updates from RSAC 2025!
- In the podcast, Jane and Rory are discussing the key insights from RSAC 2025, including the risks of model jailbreaks and harmful responses from AI systems, as highlighted by Cisco's research.
- As we delve deeper into the use of AI in cybersecurity, it's important to remember that AI security experts are emphasizing the need for ongoing evaluation of innovative models and agentic systems to combat exponentially growing attacks.
